信息安全研究 ›› 2020, Vol. 6 ›› Issue (8): 744-750.

• 技术应用 • 上一篇    下一篇

基于故障历史数据的软件可靠性测试过程 故障时间定位方法研究

孙正,尚京威,王强   

  1. 工业和信息化部电子第五研究所
  • 收稿日期:2020-08-04 出版日期:2020-08-05 发布日期:2020-08-04
  • 通讯作者: 孙正
  • 作者简介:孙 正 学士,工程师,主要研究方向为软件可靠性、软件测评。 sunzheng@ceprei.com 尚京威 硕士,工程师,主要研究方向为软件可靠性、软件质量评价。 shangjingwei@ceprei.com 王 强 硕士,高工,主要研究方向为软件质量管理、软件工程、软件可靠性工程等。 wangqiang@ ceprei.com

Research on fault time location method in software reliability testing process based on fault history data

  • Received:2020-08-04 Online:2020-08-05 Published:2020-08-04

摘要: 目前软件可靠性定量评估主要基于软件可靠性测试或真实使用中收集的故障数据进行,但是由于软件可靠性测试的测试周期很长且难以收集到足够的故障数据,限制了该项技术在实际工程中的应用;而在软件开发过程中进行测试发现的大量软件故障却由于与实际使用过程无关或没有故障时间记录而无法用于软件可靠性定量评估。本文根据软件可靠性测试剖面的特点提出软件可靠性测试输入空间模型并依据该模型生成测试用例,与常规软件测试中发现的故障输入空间进行数据匹配,对这些故障数据在软件可靠性测试中可能的故障时间进行定位,使其达到了软件可靠性定量评估的条件。并通过以某型仿发动机控制软件为实验对象,验证了方法的可行性与有效性。

关键词: 故障历史, 输入空间, 数据匹配, 故障时间定位, 定量评估

Abstract: At present, the quantitative evaluation of software reliability is mainly based on software reliability test or fault data collected in real use. However, due to the long test period of software reliability test and the difficulty in collecting enough fault data, the application of this technology in practical engineering is limited.However, a large number of software failures discovered by testing during software development cannot be used for software reliability quantitative evaluation because they are irrelevant to the actual use process or have no failure time record.In this paper, according to the characteristics of the software reliability testing profile software reliability test input space model is proposed and based on this model to generate test cases, with conventional fault found in software testing data matching the input space, the failure data in software reliability testing possible to locate fault time, to the terms of quantitative evaluation of software reliability.The feasibility and effectiveness of the method are verified by taking the control software of an imitation engine as the experimental object.

Key words: Failure history, Input space, Data matching, Fault time location, Quantitative